man3/fma.3

   1 .\" Copyright 2002 Walter Harms (walter.harms@informatik.uni-oldenburg.de)
   2 .\" Distributed under GPL, 2002-07-27 Walter Harms
   3 .\" Modified 2004-11-15, Added further text on FLT_ROUNDS
   4 .\"     as suggested by AEB and Fabian Kreutz
   5 .\"
   6 .TH FMA 3  2002-07-27 "" "Linux Programmer's Manual"
   7 .SH NAME
   8 fma, fmaf, fmal \- floating-point multiply and add
   9 .SH SYNOPSIS
  10 .nf
  11 .B #include <math.h>
  12 .sp
  13 .BI "double fma(double " x ", double " y ", double " z );
  14 .br
  15 .BI "float fmaf(float " x ", float " y ", float " z );
  16 .br
  17 .BI "long double fmal(long double " x ", long double " y ", long double " z );
  18 .fi
  19 .sp
  20 Compile with \-std=c99; link with \-lm.
  21 .SH DESCRIPTION
  22 The
  23 .BR fma ()
  24 function computes
  25 .IR x " * " y " + " z .
  26 The result is rounded according to the
  27 rounding mode determined by the value of FLT_ROUNDS.
  28 FLT_ROUNDS indicates the implementation-defined rounding
  29 behaviour for floating-point addition,
  30 and has one of the following values:
  31 .IP \-1
  32 The rounding mode is not determinable.
  33 .IP 0
  34 Rounding is towards 0.
  35 .IP 1
  36 Rounding is towards nearest number.
  37 .IP 2
  38 Rounding is towards positive infinity.
  39 .IP 3
  40 Rounding is towards negative infinity.
  41 .PP
  42 Other values represent machine-dependent, non-standard rounding modes.
  43 .SH "CONFORMING TO"
  44 C99
  45 .SH "SEE ALSO"
  46 .BR fenv (3),
  47 .BR remainder (3),
  48 .BR remquo (3)